Reducing wait time of providers of ride services using zone scoring

Abstract

A travel coordination system provides information to providers to reduce the wait time between trips. A region is partitioned into zones and generates a score for each zone. A zone score can be generated by estimating the wait time for the zone, which may be determined by a model for the wait time. The model for the wait time may use factors that contribute to the wait time, such as the number of providers in a zone and the trip request rate. The zone score for each zone is displayed to the provider on a road map of the geographic region. The travel coordination system also provides routes that use zone scoring to reduce the wait time for receiving an assignment request during travel to the destination. The travel coordination system identifies a destination and generates candidate routes to the destination. A route is selected based on route scores.

What is claimed is: 
     
         1 . A method comprising:
 receiving, at a travel coordination system from a provider device, a current location of a provider that is available to transport a rider, the provider operating the provider device in communication with the travel coordination system;   identifying, within a geographic region, a plurality of zones that are near the current location of the provider;   determining, for each zone in the plurality of zones, an estimated arrival time at the zone from the current location of the provider;   determining, for each zone in the plurality of zones, an estimated wait time at the estimated arrival time at the zone, the estimated wait time being an estimated amount of time the provider is expected to wait in the zone before receiving an assignment request to transport a rider;   generating, for each zone in the plurality of zones, a zone score based on the estimated wait time of the zone; and   causing the provider device to display a user interface of a geographical region, the user interface displaying an indication of the zone score for each zone in at least some of the plurality of zones.   
     
     
         2 . The method of  claim 1 , wherein determining an estimated wait time for a zone at the estimated arrival time for the zone comprises:
 identifying a plurality of providers located within the zone that have become available to provide a ride within the zone and received an assignment request within the zone;   determining the wait time for each provider in the plurality of providers between when the provider dropped off the rider and when the provider received the assignment request;   generating a wait time model based on the wait times for the plurality of providers, the wait time model configured to predict a wait time for the zone as a function of time; and   determining an estimated wait time for the zone at the estimated arrival time by applying the wait time model.   
     
     
         3 . The method of  claim 1 , wherein determining an estimated wait time for a zone comprises:
 determining, at a first time point, a first set of providers within the zone that are available to provide transportation to a rider;   determining, at a second time point, a second set of providers within the zone that are available to provide transportation to a rider, the second set of providers being a subset of the first set of providers, and the second time point being later than the first time point;   determining a proportion of providers within the zone that received an assignment request within a period of time between the first time point and the second time point based on the difference between the first set of providers and the second set of providers;   determining a likelihood that a provider within the zone will receive an assignment request based on the proportion of providers that received an assignment request; and   determining an estimated wait time for the zone based on the likelihood that a provider will receive an assignment request.   
     
     
         4 . The method of  claim 1 , wherein determining an estimated wait time for a zone comprises:
 generating a model of provider supply for the zone, the model of provider supply estimating a quantity of providers in the zone that are available to transport a rider;   generating a model of provider demand estimating a quantity of riders in the zone requesting transportation; and   determining an estimated wait time for the zone based on the model of provider supply and the model for provider demand.   
     
     
         5 . The method of  claim 1 , wherein the zone scores are adjusted based on a likelihood that the provider will travel to a zone based on a zone score. 
     
     
         6 . The method of  claim 1 , wherein the zone scores are adjusted based on a preference the provider has shown for a zone. 
     
     
         7 . The method of  claim 1 , wherein the estimated wait time in a zone is modeled using an exponential distribution. 
     
     
         8 . The method of  claim 1 , wherein the generated zone scores are cached by the travel coordination system. 
     
     
         9 . The method of  claim 1 , wherein the provider device is a personal computing device running a client application for the travel coordination system. 
     
     
         10 . A method comprising:
 receiving, at a travel coordination system from a provider device, a current location of a provider that is available to transport a rider, the provider operating the provider device in communication with the travel coordination system;   identifying, within a geographic region, a plurality of zones that are near the current location of the provider;   determining, for each zone in the plurality of zones, an estimated arrival time at the zone from the current location of the provider;   generating, for each zone in the plurality of zones, a provider supply model for the zone, the provider supply model estimating a quantity of providers in the zone that are available to transport a rider;   generating, for each zone in the plurality of zones, a provider demand model estimating a quantity of riders in the zone;   estimating, for each zone in the plurality of zones, provider supply and provider demand in the zone at the arrival time based on the provider supply model and the provider demand model;   generating, for each zone in the plurality of zones, a zone score based on the estimated provider supply and estimated provider demand, the zone score representing an estimated wait time for the provider in the zone before receiving an assignment request to transport a rider; and   transmitting data corresponding to the zone scores for the plurality of zones to the provider device.   
     
     
         11 . The method of  claim 10 , wherein the zone scores are adjusted based on a likelihood that the provider will travel to a zone based on a zone score. 
     
     
         12 . The method of  claim 10 , wherein the zone scores are adjusted based on a preference the provider has shown for a zone. 
     
     
         13 . The method of  claim 10 , wherein the generated zone scores are cached by the travel coordination system. 
     
     
         14 . The method of  claim 10 , wherein the model of provider supply is based on at least one of: a number of providers in the zone, a rate at which providers begin to accept assignment requests, a rate at which providers stop accepting assignment requests, a rate at which providers are being assigned to riders, a rate at which providers are completing trips in the zone, a rate at which providers are traveling into and out of the zone, and a historical amount of provider supply in the zone. 
     
     
         15 . The method of  claim 10 , wherein the model of provider demand is generated based on at least one of: a trip request rate, an assignment request rate, and a historical amount of demand in the zone. 
     
     
         16 . The method of  claim 10 , wherein the provider is a person driving a car, bus, truck, boat, or other motorized vehicle capable of transporting passengers. 
     
     
         17 . The method of  claim 10 , wherein the provider is an autonomous vehicle that receives routing instructions from the travel coordination system. 
     
     
         18 . The method of  claim 10 , wherein generating a model for provider supply for a zone comprises using data from zones surrounding the zone in a k-ring. 
     
     
         19 . The method of  claim 10 , wherein generating a model for provider demand for a zone comprises using data from zones surrounding the zone in a k-ring. 
     
     
         20 . A non-transitory computer-readable storage medium storing computer program instructions to perform steps comprising:
 receiving, at a travel coordination system from a provider device, a current location of a provider that is available to transport a rider, the provider operating the provider device in communication with the travel coordination system;   identifying, within a geographic region, a plurality of zones that are near the current location of the provider;   determining, for each zone in the plurality of zones, an estimated arrival time at the zone from the current location of the provider;   determining, for each zone in the plurality of zones, an estimated wait time at the estimated arrival time at the zone, the estimated wait time being an estimated amount of time the provider is expected to wait in the zone before receiving an assignment request to transport a rider;   generating, for each zone in the plurality of zones, a zone score based on the estimated wait time of the zone; and   causing the provider device to display a user interface of a geographical region, the user interface displaying an indication of the zone score for each zone in at least some of the plurality of zones.